The invention relates to a redundancy manipulator repetitive motion planning method based on a Taylor type discrete periodic rhythm neural network under periodic noise. The redundancy manipulator repetitive motion planning method includes the steps that (1), inverse kinematics of a manipulator on an angular acceleration layer is analyzed by using quadratic optimization, and a jacobian equation forangular acceleration is designed; (2), an angular velocity norm is equivalent to an angular acceleration norm, and a standard quadratic planning program is established; (3), a periodic rhythm neuralnetwork is discretized through a Taylor differential method, a solver of the Taylor type discrete periodic rhythm neural network is obtained and used for solving the standard quadratic planning program; and (4), a solving result drives each joint motor through a controller so that the manipulator can conduct repetitive motion planning from the initial position. According to the redundancy manipulator repetitive motion planning method based on the Taylor type discrete periodic rhythm neural network under the periodic noise, the actual track of the manipulator coincides with the desired path, and repetitive motion planning of the manipulator can be achieved; and the periodic rhythm neural network is successfully discretized and thus is easier to achieve in hardware, and the periodic noise issuppressed in the hardware.